The prediction of the three dimensional structure of a globular protein from its amino acid sequence is an extremely important and unsolved problem of contemporary molecular biology. The overall objective of the proposed FIRCA project, which is an extension o an NIH proposal GM-48835 entitled "Interaction Pattern Descriptor of Protein Structure is the development of a methodology to predict the tertiary structure of globular proteins. The basic idea is to construct a library of possible protein folds. Subsequently, the sequence of interest is threaded through all such folds and the structure with the lowest energy is assigned to be the putative native state. To achieve this objective, the following approach will be taken in this proposal: (1) Using a new high coordination lattice, each protein fold will be represented at the level of alpha-carbons. The existing library of anti parallel beta proteins will be extended to include parallel beta, mixed motif alpha/beta, an alpha plus beta and helical folds. (2) The extant self-consistent potential of mean force which includes hydrophobic interactions and hydrogen bonding, will be extended to include both van der Waals and electrostatic interactions. (3) The methodology will tested by screening the structural library against the Swiss Prot sequence database. To fully successful, it must identify all sequences whose structure is known and found in the idealized structural library. (4) The tertiary structure of a number of proteins whose structure is currently being determined by experiment will be predicted. (5) Using knowledge-based rules for the arrangement and handedness of supersecondary structure elements, e.g. beta-alpha-beta connections must be right-handed, a library of all possible six seven, and eight-stranded anti-parallel beta proteins and all mixed alpha/beta protein containing from six to eight helices will be constructed. This will provide a library of folds, not all of which have been experimentally observed. The sequence database will be screened against this extended library to provide a compilation of predicted structures.